The Agent Skills Directory

[SAFE]: The skill instructions and configuration files exclusively address Git repository management through rebasing. No malicious patterns such as remote code execution, obfuscation, or data exfiltration were detected.
[SAFE]: The logic explicitly requires the agent to preserve local work and avoid committing or stashing dirty files without authority, which is a security best practice for data integrity.
[SAFE]: The evaluation triggers and scenarios focus on legitimate developer workflows (e.g., handling detached worktrees, resolving mechanical conflicts in lockfiles, and identifying remote default branches).

pull